Your goal in a technical interview is to impress the interviewer. At large companies the only way this happens is by knowing the problem as good as, or better than the interviewer.

First get all the basics down, algorithms and data types are almost always going to be more needed and useful than dynamic programming questions. So study things like stacks queues hashset/hashmap BFS DFS etc.
Once you have a good grasp on basics start practicing a lot of leetcode. Work out what the problem is, and figure out how to characterize similar problems together. If you're struggling to do this on your own you should reference other people until you get the idea (I liked a lot of neetcode's work on YouTube, worked well for my adhd).

You also want to be able to envision different possible solutions. For this you should have a good understanding of big O notation as well as space and time complexity. Most problems have solutions that are optimal in one but not the other (ex. Most hashmap problems are efficient in time, O(1) lookup, but not in space O(N).). You should ask your interviewer what kind of solution they want, and then be able to provide a solution of that type (ask if they care more about time or space. Maybe about how big your array of values is, or how much memory you have, etc).

If you have people to practice with you should do it. You will typically learn and improve solutions by talking about them with others. Some bigger cities have meetups for this idea.

Also make sure you know basic information about whatever language you plan to test in. It's really common to be asked why you use the language, or what it's advantages are.

Let me know if you have any questions. Good luck on your interviews, and study hard!

Should note this is going to be mostly for lower level technical interview ^